Model versions and quantization

Why the same model comes in several files (fp16, fp8, GGUF), how to tell which fits your graphics card, and what it costs in quality.

Open the download page for almost any local model and you will find not one file but several -- names ending in fp16, fp8, or GGUF Q8, Q4, and so on. They are all the same model. What differs is precision: how much detail is kept in each of the model's numbers. Lower precision means a smaller file that needs less graphics memory and often runs faster, at some cost to quality. This is called quantization.

The practical upshot: a model that "needs a 24 GB card" at full precision will often run happily on a 12 GB card in a smaller version. Picking the version that fits your card is the key skill for running models locally.

One warning about the word "version", which this page uses in a narrow sense. Here it always means which build of one model -- fp16 against fp8 against GGUF. It does not mean which release, as in Wan 2.1 against Wan 2.2: those are two different models, with two different names. See Models, builds and LoRAs.

The builds, from biggest to smallest

  • fp16 / bf16 (full precision) -- the original quality, the largest file, and the most graphics memory. Use it if your card has room.
  • fp8 (half precision) -- roughly half the memory of fp16 for a small, usually hard-to-notice drop in quality. A good default when the full build is too big.
  • GGUF (Q8 down to Q4 and lower) -- a family of compressed versions that let big models run on modest cards. Q8 is close to full quality; the quality cost grows as the number drops, and heavy compression (Q4 and below) can visibly soften results. GGUF files need a GGUF-aware setup to load them.

Which version fits my card?

These are rough starting points, not guarantees -- actual fit depends on the specific model, the resolution you generate at, and what else is loaded:

Graphics memory Rough guidance
8 GB Smaller models, or GGUF (Q4-Q6) of larger ones. Video will be a stretch -- consider the cloud.
12 GB fp8 of most image models; GGUF for the largest ones. Short video is possible for smaller models.
16 GB fp8 comfortably; fp16 for many image models.
24 GB and up fp16 of most things, including video.

If your card is below the model you want, you have two easy options: use a smaller version, or run that model on Comfy Cloud instead.

A build never changes the look

Every build of a model makes the same kind of picture. A lighter one fits a smaller card and gives up a little detail; it does not change the style, the composition or what the model is good at. That is why it is safe to treat this as a hardware question and nothing more.

You usually do not choose by hand

You pick the build once, when you first choose that model for an operation, and it is described to you by what it is -- its name, its download size, the memory it wants -- rather than by a filename, so you rarely have to think in terms of fp8 or Q4 at all. From then on that is the build that runs, locally and on Comfy Cloud alike; it is never swapped for a smaller one behind your back, because that would change the picture. The other time you think about builds directly is when you already have a particular file on disk and want to use it -- which the Files needed list, in the ComfyUI section of Settings, lets you do.

Quantization methods and the quality they deliver keep improving. Treat the specifics here as a current snapshot and check a model's own notes for anything it recommends.